All 113 E. coli strains used in this study are listed in Supplementary Table S1. Among them, 104 E. coli strains isolated from the brains of ducks with septicemia and neurological clinical symptoms in Eastern China in previous study, were included (Chen et al. 2017). The remaining 9 strains were selected from the collection in our laboratory. These E. coli strains were used to determine the host range of phages. Phages WG01 and QL01 were used as the parental phages for homologous recombination, while E. coli strains DE017 and DE205B were used as host strains of the two phages for amplification, respectively (Chen et al. 2017).
Luria-Bertani (LB) liquid medium were used to cultivate bacterial strains. When necessary, 100 μg/mL ampi- cillin was added. LB 0.5% soft agar plates and LB 1.5% agar plates were used for phages plaque assay. Tryptic soy broth liquid medium (TSB, Merck, Germany) were used to culture E. coli biofilm. SM buffer (50 mmol/L Tris-HCl [pH 7.5], 100 mmol/L NaCl, 10 mmol/L MgSO4 and 0.01% gelatin) was used for the dilution and storage of the phages. Phage cultures were clarified by centrifugation at 5000 ×g, 4 ℃ for 10 min, then the supernatants were filtered through 0.22 μm filters (Merck Millipore, Germany) and stored at 4 ℃. Phosphate Buffer Saline (0.1 mol/L Na2HPO4, 0.15 mol/L NaCl2, pH 7.2) was used to wash and dilute E. coli strains.
We selected six diverse DNA fragments in gene 37 from QL01 as the recombination regions to replace the corresponding fragments in WG01 (Bartual et al. 2010; Chen et al. 2017). Six different fragments of gene 37 (i.e., QLae, QLbe, QLce, QLde, QLfe and QLm) were amplified (Fig. 1) and fused using the primers listed in Supplementary Table S2, with WG01 and QL01 phages as the DNA templates. Plasmids construction was performed as described previously (Yoichi et al. 2005). In brief, the PCR products were digested with BamH I and EcoR I, followed by insertion into plasmid vector pUC118 (Takara, Shiga, Japan), generating six different recombinant plasmids, pUC118ae, pUC118be, pUC118ce, pUC118de, pUC118fe and pUC118m.
Figure 1. Sketch map of gene 37 in QL01 and the results of the isolated chimera phages. Different color fragments from QL01 for WG01 recombination. Four types of recombinant phages, namely, WGqlae, WGqlbe, WGqlce and WGqlde, were isolated from DE205B, the host of QL01. The symbol X indicate that no chimeric phages were isolated with DE205B by infecting the plasmid-containing cell with phage WG01.
Homologous recombination of T4-like phages was performed as described previously (Chen et al. 2017; Yoichi et al. 2005) with minor modifications. Briefly, competent strain DE017 were transformed with the recombinant plasmid, pUC118ae, pUC118be, pUC118ce, pUC118de, pUC118fe and pUC118m, respectively. The DE017 transformant strains were cultured in 5 mL LB liquid medium with 50 ig/mL ampicillin at 37 ℃ until OD600 0.4, then the phage WG01 was added. After purification and fil- teration, recombinant phages were isolated using E. coli DE205B (susceptible to QL01, but resistant to WG01) as host bacteria by double-layer plate method (Adams 1959). Recombinant phages were purified by plaque assay. PCR and Sanger sequencing analysis of gene 37 from the recombinant phages was carried out using primers Fcheck/ Rcheck (Chen et al. 2017).
The recombinant phages, WGqlae, WGqlTbe, WGqlTce, WGqlde, and the parent phages WG01 and QL01 were prepared by double-layer plate method (Adams 1959). A total of 113 avian E. coli strains were used as the host bacteria to determine the host ranges (Supplementary Table S1) using spot test with minor modifications (Chen et al. 2016). In brief, bacterial cultures at exponential phase (100 μL) were spread on an LB agar plates. Each of six phage suspensions (10 μL at 109 PFU/mL) was spotted on the surface of the bacteria lawn. The plates were examined for lysis after 8 h incubation at 37 ℃.
Thermo and pH stability of phages was evaluated using a previously described protocol with minor modifications (Laemmli 1970). Phage suspensions (100 μL) were mixed with SM buffer (900 μL) in 1.5 mL sterile Eppendorf tubes, and were kept at different temperatures (i.e. 30-60 ℃) for 1 h. The survival rates of phages were determined by double-layer method as described above.
To determine the pH stability of phages, 100 μL phage suspension was added to 900 μL SM buffer with a gradient of pH values (range from 4 to 11), followed by incubation at 37 ℃ for 1 h. Similarly, double-layer plating method was used to examine the survival rates of phages.
Multiplicity of infection (MOI) was defined as the ratio of viral particles to potential host cells. Optimal MOIs of phage WG01, QL01 and WGqlae were determined by using standard protocol with minor modifications (Lu et al. 2003). Phages were diluted to 104-109 PFU/mL. Log phase host bacteria were adjusted to 1 × 108 CFU/mL (Table 1). Phages at different titer (100 μL, MOI = 0.0001, 0.001, 0.01, 0.1, 1, 10) and 100 μL 1 × 108 CFU/mL log phase host bacteria were added to 1.5 mL sterile Eppendorf tubes containing 800 μL LB medium, followed by incaution at 37 ℃ for 3.5 h with shaking (60 rpm). Bacteria cultures (phage-free) and phage cultures (cell-free) were used as control groups in this experiments. All assays were conducted in duplicate. The MOI that generated the highest phage titer within 3.5 h was considered as optimal MOI.
MOI Phages (PFU/mL) Bacteria (CFU/mL) Bacteriophage titer after 3.5 h (PFU/mL) Proliferation times WG01 QL01 WGqlae WG01 QL01 WGqlae 0.0001 1 × 103 1 × 107 2.82 × 105 3.27 × 105 1.77 × 106 282 327 1770 0.001 1 × 104 1 × 107 3.86 × 106 4.64 × 107 1.53 × 106 386 4640 153 0.01 1 × 105 1 × 107 5.2 × 106 2.47 × 107 1.31 × 107 52 247 131 0.1 1 × 106 1 × 107 3.57 × 107 4.58 × 107 2.21 × 107 35.7 45.8 22.1 1 1 × 107 1 × 107 2.87 × 107 3.26 × 107 3.29 × 107 2.87 3.26 3.29 10 1 × 108 1 × 107 2.4 × 107 2.03 × 107 2.91 × 107 0.24 0.20 0.29
Table 1. Comparison of MOIs between recombinant WGqlae and parental phage WG01 and QL01.
Latent time and burst size of phage WGqlae were determined by one-step growth curve as previous described (Pajunen et al. 2000). 10 mL Mid-exponential phase E. coli DE205B culture (2 × 108CFU/mL, OD600 = 0.4) were mixed with 10 mL 1 × 107 PFU/mL bacteriophage suspensions. Allow the phages to specifically adsorb to the host bacteria for 10 min at 37 ℃. Then, after removing unabsorbed phages by centrifuging at 5000 rpm for 10 min, the precipitated pellet was resuspended in 20 mL fresh LB medium, followed by incubation at 37 ℃. Samples were taken out at 10-min intervals for 100 min, and 1% chloroform (final concentration) was added to release the intracellular phage for phage titration using the doublelayer agar plate method.
To evaluate the genetic stability of recombinant phage WGqlae, the 1st, 10th and 20th generation WGqlae were collected. Gene 37 was amplified by primers Fcheck/ Rcheck (Chen et al. 2017). PCR products were sequenced by Suzhou genewiz biotechnology Co., Ltd. (Suzhou, China).
Lytic activity of phages WGqlae was evaluated by OD600nm values of the host bacteria every 2 h at the MOI of 0.1, 1 and 10 using 96-well microtiter plates. Each of phage group was comprised of 100 μL host bacteria (2 × 108 CFU/mL, DE205B or DE192) and 100 μL diluted phage lysate (2 × 107-2 × 109 PFU/mL), and the assay was conducted in quadruplicates. Control group was consisted of equal volume of host bacteria suspension and fresh LB liquid medium. The 96-well microtiter plates were incubated at 37 ℃ for 24 h, and the OD600 values were measured at 2 h intervals using an ELISA microplate reader (Biotek, VT, USA).
Quantitative detection of bacterial biofilm was performed as described previously with minor modifications (Stepa-novic et al. 2000, 2004). Briefly, E. coli DE192 and DE205B were incubated in TSB liquid medium containing 1%, 2% and 3% glucose at 37 ℃ for 24 h, 48 h, 72 h, 96 h and 120 h, respectively. Supernatants were then slowly decanted, and washed three times with 200 μL PBS. After air drying, each well was added 200 μL methanol to fix for 15 min. After discarding methanol and air drying, each well was added 200 μL 1% crystal violet and incubated for 5 min. After being washed three times with PBS, each well was added 100 μL acetic acid (33% v/v) to dissolve the crystal violet, and was placed in 37 ℃ incubator for 30 min. Subsequently, the OD595nm value was measured by a microplate reader.
Fresh host bacteria DE192 and DE205B (100 μL, 2 × 108 CFU/mL) were added to 96 well plates. Before culture, WGqlae group was added 100 μL phage WGqlae (2 × 108 PFU/mL); polymyxin B group was added 100 μL polymyxin B (4 mg/L); and WGqlae + Polymyxin B group was added 50 μL phages (4 × 108 PFU/mL) and 50 μL polymyxin B (8 mg/L, minimum inhibitory concentration was 2 mg/L for DE192 and DE205B). Control group were added equivoluminal TSB medium. Culture condition and quantitative detection of bacterial biofilm were performed as mentioned above.
Escherichia coli DE192 and DE205B were incubated to form biofilm in TSB liquid medium at 37 ℃ for 48 h and 72 h, respectively. Then, WGqlae group was added 200 μL phage WGqlae (1 × 108 PFU/mL); polymyxin B group was added 200 μL polymyxin B (1 × 108 PFU/mL), and WGqlae + Polymyxin B group was added 100 μL phages (2 × 108 PFU/mL) and 100 μL polymyxin B (8 mg/L). Control group was added equivoluminal TSB medium. Culture condition and quantitative detection of bacterial biofilm were performed as mentioned above.
All statistical analyses in this study were carried out with the GraphPad Prism 5 software package. Mean difference of control and treat groups of planktonic or biofilm assay were analyzed by student's t test. P value ≤ 0.05 was considered statistically significant.
Bacteria, Bacteriophages and Growth Conditions
Construction of the Plasmids with Diverse Recombination Regions
Homologous Recombination and Isolation of Recombinant Phages
Host Ranges of Recombinant Phages
Thermo and pH Stability of WG01 Derivatives
Optimal Multiplicity of Infection
One Step Growth Curve of Recombinant Phage WGqlae
Genetic Stability of Recombinant Phage WGqlae
Lytic Capacity of Phages WGqlae against E. coli in Planktonic Form
Quantification of Bacterial Biofilm
Inhibiting Efficacy of Phage WGqlae, Polymyxin B, and Their Combination on Biofilm Formation
Efficacy of Phage WGqlae, Polymyxin B and Their Combination on the Clearance of Formed Biofilm
We randomly selected six different DNA fragments, QLae (709-2775 nt), QLbe (949-2775 nt), QLce (1189-2775 nt), QLde (1429-2775 nt), QLfe (1669-2775 nt) and QLm (1363-2383 nt) of gene 37 in QL01 to replace the corresponding fragments of WG01 using a gene manipulation experiment (Fig. 1).
Four recombinant phages (WGqlce, WGqlbe, WGqlae and WGqlde) were isolated from clear plaques on doublelayer plates with E. coli DE205B, one of host bacteria of phage QL01 (but not the host of WG01) (Fig. 2). Sequencing analysis of gene 37 confirmed the generation of chimeric phages WGqlce, WGqlbe, WGqlae and WGqlde. The gene 37 of WGqlce did not mutate in the process of homologous recombination. However, nucleic acid mutations occurred in gene 37 of WGqlae, WGqlTbe and WGqlTde, causing 14, 29 and 1 amino acid substitutions, respectively (Supplementary material S3).
Figure 2. Plaque morphology of recombination phages. Phages WGqlae, WGqlbe, WGqlce and WGqlde were grown on E. coli DE205B in double agar LB plates for 6 h. WGqlce and WGqlbe formed clear and round plaques approximately 0.1 cm in diameter. WGqlae formed clear and round plaques approximately 0.3 cm in diameter. WGqlde formed clear and round plaques approximately 0.2 cm in diameter.
Host ranges of the phages WG01, QL01 and four WG01 derivatives were determined against 113 E. coli strains. Phages WG01 and QL01 lysed 20 (17.70%) and 50 (44.25%) of the 113 E. coli strains, respectively. While phages derivatives WGqlae (with AA 237-925 of QL01 gp37), WGqlbe (with AA 317-925 of QL01 gp37), WGqlce (with AA 397-925 of QL01 gp37) and WGqlde (with AA 477-925 of QL01 gp37) lysed 49 (43.36%), 52(46.02%), 52 (46.02%) and 52 (46.02%) of the 113 E. coli strains, respectively (Fig. 3, Supplementary Table S1). Among them, WGqlde, which was only replaced the downstream 1429-2772 bp in gene 37 of QL01, can lyse the most part of host strains of QL01. This result suggested that this domain of gp37 mainly determined the host range, which is consistent with the finding that C-terminal domain determined the receptor specificity. WGqlae, however, can infect four additional host bacteria, DE011, DE037, DE061 and DE077, in comparison to the parental phages, which suggested the expansion of host-range. Therefore, the recombination phage WGqlae was chosen to analyze the basic characteristic and antibacterial effect.
Figure 3. Host ranges analysis of recombinant and their parental phages. Phages derivatives WGqlae, Wgqlbe, Wgqlce and Wgqlde lysed 49, 52, 52 and 52 strains of the 113 E. coli strains, respectively (Only 64 strains were showed in this map, because other 47 strains were not susceptible to any of these phages). WGqlae can infect 4 additional host bacteria DE011, DE037, DE061 and DE077 in comparison to the parental phages.
Thermo stability test was carried out at pH 7 to investigate thermo stability of phages. The results showed that more than 60% of phage WGqlae remained active after incubation at 30 ℃, 40 ℃ and 45 ℃ for 1 h, which were similar to the results from WG01 and QL01. In contrast, less than 2% of three active phage particles were observed at 60 ℃ (Fig. 4A). The results showed that the thermo stability of the WG01 derivative WGqlae was nearly the same as donor phage QL01.
Figure 4. Biological properties of the recombinant phage WGqlae and its parental phages. A Thermo stability of recombinant phage WGqlae and the parental phages WG01 and QL01. B The pH stability of recombinant phage WGqlae and its parental phages WG01 and QL01. C One-step growth curve of the recombinant phage WGqlae and its parental phages WG01 and QL01. All the values are the means of 2 determinations.
The pH stability of phages was tested at 37 ℃ for 1 h. Phage WGqlae, WG01 and QL01 showed similar pH stability. In detail, they all showed the highest stability at pH 7, and they demonstrated similar survival rate (> 50%) at pH 5-9 (Fig. 4B). For all three phages, the phage titers decreased sharply at the acidic (pH = 4) or alkaline (pH = 10, 11) conditions.
MOI was defined as the ratio of viral particles to potential host cells. The optimal MOI of parental phage WG01 and QL01 were both 0.001. After incubation for 3.5 h, phage WG01 and QL01 multiplied 386 times and 4640 times, respectively. While the optimal MOI of phage WGqlae was 0.0001, and WGqlae increased by 1770 times after 3.5 h proliferation. The optimal MOI of recombinant phages WGqlae was lower than those of parental phage WG01 and QL01 (Table 2).
Time Control DE192 (OD595) DE205B (OD595) LB TSB 1% G-TSB 2% G-TSB 3% G-TSB LB TSB 1%G-TSB 2%G-TSB 3%G-TSB 24 h 0.29 0.49 0.53 0.53 0.53 0.45 0.63 0.41 0.46 0.38 0.34 48 h 0.34 1.06 3.81 2.02 1.95 1.58 1.76 1.57 1.50 1.18 1.34 72 h 0.32 1.43 2.38 2.37 2.38 1.59 1.37 1.86 1.14 0.90 1.04 96 h 0.31 0.63 0.82 1.06 1.02 1.05 0.92 0.67 0.67 0.68 0.61 120 h 0.33 0.78 1.16 1.47 1.24 1.25 1.21 0.89 0.74 0.78 0.88
Table 2. Optimal biofilm formation conditions of DE192 and DE205B.
One-step growth experiments were performed to compare latent time period and burst size of the phage WGqlae and the parental phages. One-step growth curve consisted of three phases-latent, log and stationary phases. Latent time period of the phage WGqlae was 10 min, and the burst size was about 110 phages/cell (Fig. 4C). In comtrast, phage WGqlae showed similar to phage QL01 in one-step growth experiments.
The gene 37 of phage WGqlae in 1st, 10th and 20th generation did not showed any mutations, suggesting that recombinant phage WGqlae was genetically stable.
The OD600 values of the phage groups were significantly lower than that of the control group in 24 h, indicating that the derivative WGqlae inhibited bacteria proliferation. The OD600 values of the control group of both host bacteria DE192 and DE205B increased continuously in 24 h. All OD600 values of three phage groups (MOI 10, 1 and 0.1) began to decrease after a short increase in the first 2 h, and it did not increase until 16 h. In the first increased phase, the smaller MOI values were correlated with the more obvious increased trend, but in the second increased phage, the larger MOI values were associated with the more obvious decreased trend (Fig. 5).
Our results showed that optimal medium for biofilm formation of host bacteria DE192 and DE205B was TSB medium without glucose, and optimal culture time of DE192 and DE205B was 48 h and 72 h, respectively. Therefore, the optimal medium and conditions were subsequently used in the following biofilms test.
Recombinant phage WGqlae (P < 0.01), polymyxin B (P < 0.05) and the mixture of WGqlae + Polymyxin B (P < 0.01) inhibited the biofilm formation of host bacteria DE192 (Fig. 6A) and DE205B (Fig. 6B), significantly. When phage WGqlae was used alone, the biofilm of host bacteria DE192 and DE205B decreased by 38.59% and 38.20% respectively (P < 0.01), compared with the control group. As shown in Fig. 6A, phage WGqlae had better antibacterial effect than antibiotic Polymyxin B (P < 0.05). As shown in Fig. 6B, when the mixture of WGqlae + Polymyxin B was used, Polymyxin B significantly enhanced the ability of phage WGqlae to inhibit biofilm formation of strain DE205B (P < 0.01). Comparing with the control group, the biofilm formation of host bacteria DE192 and DE205B in WGqlae + Polymyxin B group was reduced by 40.94% (P < 0.01) and 77.53% (P < 0.001), respectively (Fig. 6).
Comparing with the control group, the formed mature biofilm of E.coli DE192 and DE205B in phage WGqlae group decreased by 32.23% (Fig. 7A) and 33.22% (Fig. 7B), respectively (P < 0.01). No significant difference was observed between Polymyxin B group and the control group, suggesting Polymyxin B had no effect on established mature biofilm (both Fig. 7A, 7B). There was also no significant difference between WGqlae + Polymyxin B group and the only phage WGqlae group in removing the formed biofilm, which suggested that Polymyxin B can contribute to inhibiting biofilms formation but not to remove the formed mature biofilm of DE192 (Fig. 7A) and DE205B (Fig. 7B).